I found this recent viral article to be really sensationalized. AlphaZero makes a better chess playing algorithm. Media's reaction is "Oh my gosh, general AI is here. Everybody run for cover! The robots are coming!" Then the scientists don't bother correcting them and actually encourage it using clever choices of words. Tempering expectations doesn't help VC funding after all.
Just because an "AI" does one task well with highly structured and discrete move possibilities, doesn't mean it can do any task well much less in the real world. The AI scientists in the 1960s promoted the same hype when an algorithm could beat humans at checkers.
https://news.yahoo.com/deepmind-apos-alphazero-now-showing-190000147.html